Case Studies. Framework calibration + landmark AI infra retrospectives
Eleven case studies across two shapes: three calibrating the framework's 2×2 verdict on live 2024-2026 deals, and eight retrospectives on landmark AI infrastructure decisions. Each maps the timeline, the strategic thesis, what actually happened, and the lessons for practitioners now specifying similar transactions. Reference material for PE sponsors, corp-dev teams, and strategy leaders.
01Framework calibration case studies
Three live-deal studies calibrating the framework's 2×2 output (UNDERWRITE / STRUCTURE / RESTRUCTURE / PASS) and the uncontracted-principal veto (GREEN / AMBER / RED). Each shows per-layer scoring, verdict pill, and the assessment basis for what is public [P], derived [D], and illustrative [I].
-
AAirTrunk · Blackstone + CPP · UNDERWRITE benchmarkA$24bn Sep 2024 · the clean-deal calibration anchor. Both scores LOW, veto GREEN.
-
BCoreWeave $2.6bn DDTL 3.0 · the AMBER worked exampleJul 2025 SEC crwv-20250728 · parent-guarantee absorbs an open tail that would fire the veto in a thin-equity SPV.
-
CCrusoe / Stargate Abilene · RESTRUCTURE, realizedFramework flagged single-tenant concentration pre-close. Oracle + OpenAI dropped in Mar 2026; Microsoft + Meta stepped in.
02The retrospective essays
-
01Blackstone/QTS five years onReference LBO of the AI era. $21B EV take-private
-
02Microsoft/Constellation TMI two years inHyperscaler nuclear PPA. First large restart
-
03Amazon/Talen: FERC BTM rejectionCautionary case on behind-the-meter arrangements
-
04CoreWeave. 18 months post-IPOAI-native GPU cloud public-market reference
-
05Iron Mountain data-centre pivot10-year adjacent-market entry + investor rerating
-
06Digital Realty/Teraco Africa four years onEmerging-market DC platform via M&A
-
07FTC challenge to Nvidia/ARMAI-adjacent M&A antitrust reference
-
08CMA UK hyperscaler cloud investigationCloud market antitrust precedent + remedies
Working on an AI infrastructure decision?
I advise PE sponsors + corporate strategy teams on transaction structuring, technical DD, and regulatory sequencing for AI infrastructure deals of similar shape.
How I work →